Taxonomy Import - 1 Preparing Data

Interest Manager – Importing Topic Types & Topics (Part 1: Preparing Your Data)

Audience: Business admins, implementation consultants
Next: [Part 2 – Running the Import]

What this does

Interest Manager (IM) organizes interests into Topic Types (e.g. Country, Sector) and the
Topics under them (e.g. United States, Energy). Topics can be bulk-loaded from a spreadsheet
instead of being created one at a time.

 Your CSV  ──►  Topic Mapping (staging)  ──►  Run Topic Import  ──►  Topic Types + Topics
                 Status = Pending                                      Mapping rows → Imported / Error

The import is safe to re-run. It never creates a duplicate Topic Type or Topic. Rows that match
existing records are linked to them.

Where the data comes from

Pick whichever source fits:

1. Build the CSV yourself (most common). Use a spreadsheet with the columns below.
2. Generate it from an RiXML research file. IM can read a sample RiXML file and produce a draft
   CSV with every tag and value already filled in. You then delete the rows you don't need and tidy
   the names. (Your technical admin runs this step; see Part 2, Step 1.)

CSV columns

* Only rows with Status__c = Pending are processed.

Leave all other Topic Mapping fields empty. The import fills them in.

How duplicates are handled

Topic Types: one per name

Every row with the same Topic_Type_Name__c goes to the same Topic Type. Spell the names exactly
the same way, including capitalization and spacing.

Topics: one per Type + Name (the default)

By default IM treats every Topic Type as canonical: one Topic per Topic Type + Topic Name,
wherever the value came from. That lets many sources share one clean list.

Result: 2 Topic Types (Country, Region) and 2 Topics (United States, North America).

Tip: Standardize names across the CRM. Use United States every time, not a mix of US,
U.S. and USA.

Exception: non-canonical types (rare)

Some types need one Topic per source tag, even when two tags share a display name. Your technical
admin lists those types in the AFR setting ACE.FPM.Settings.NonCanonicalTypeNames. For these
types, rows must include Tag_Path__c and Tag_Value__c, and each distinct path + value becomes its
own Topic. If you are unsure, leave this setting alone.

Parent / child topics (optional)

To nest topics (e.g. North America under Americas):

1. Give every row involved an External_Id__c.
2. On the child row, put the parent's ID in Parent_Ids__c. Separate multiple parents with commas.

Best practice: load parents and children in the same file. If a parent is loaded later, your
technical admin can rebuild the links afterwards (Part 2, Step 5).

Sample CSV

Topic_Type_Name__c,Topic_Name__c,Status__c,Comments__c,External_Id__c,Parent_Ids__c
Asset Class,Equity,Pending,,AC-EQ,
Asset Class,Fixed Income,Pending,,AC-FI,
Region,Americas,Pending,,RGN-AMER,
Region,North America,Pending,US and Canada,RGN-NA,RGN-AMER
Country,United States,Pending,,CTY-US,RGN-NA
Sector,Energy,,Load later,SEC-EN,

Result: 4 Topic Types and 5 Topics, plus 2 parent/child links. The Energy row is loaded but
skipped until its Status is changed to Pending.

Checklist before handing off

• Each row has Topic_Type_Name__c and Topic_Name__c (80 characters or fewer).
• Topic Type Names are spelled the same way everywhere.
• Topic Names use standard CRM wording.
• Rows to import have Status__c = Pending.
• RiXML Tag_* columns are untouched.
• If you use parent/child links, every Parent_Ids__c value matches an External_Id__c.
• Saved as CSV (UTF-8).